New Posts New Posts RSS Feed - 1970s Design Indulgence
  FAQ FAQ  Forum Search   Events   Register Register  Login Login

Welcome to the Graham Slee Audio Products Owners Forum

 

Open to all owners plus those contemplating the purchase of a Graham Slee HiFi System Components audio product and wishing to use this forum's loaner program: join here (Rules on posting can be found here)

This website along with trade marks Graham Slee and HiFi System Components are owned by Cadman Enterprises Ltd


1970s Design Indulgence

 Post Reply Post Reply Page  <1 254255256257258 345>
Author
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Retired

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 16298
Post Options Post Options   Thanks (1) Thanks(1)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 27 Nov 2022 at 7:40am
How about signal to noise?


Above is unweighted r.m.s. noise, 20Hz - 20kHz


Above is CCIR 1k weighted noise, 20Hz - 20kHz


And this is A weighted - often used because it looks best - 20Hz - 20kHz
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps
Back to Top
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Retired

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 16298
Post Options Post Options   Thanks (2) Thanks(2)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 27 Nov 2022 at 8:19am
And distortion?


At high frequencies, noise is interfering with the distortion readings, so the curve isn't smooth like you'd get from line-level. In our most sensitive hearing region, it is better than 0.01% THD.


Above is the readout for 1kHz. The near 0.01% THD includes noise (THD+N). And as the RIAA response applies more gain at low frequencies, some of that noise has been filtered out by the high pass filter which is shown set to 160Hz.


Above is the distortion at +23dB input, that's an input of 70mV (ref 1kHz) which will never happen unless the cartridge is a really high output (14mV) and on extreme transients which are +14dB. The "S/NAD" some like looking at is shown inset, but the FFT was set a little too fast which caused the 1kHz fundamental to be displayed a little bit lower than +23dB. But the HF noise right up to 1MHz has been captured realistically, and is better than -100dB.


Above shows the unclipped sine wave at +23dB. The THD is still acceptable at better than 0.04%.


And again but giving the readout of the maximum output voltage at +23dB in. The 'scope inset shows approx. +/- 8V and the r.m.s, is 0.707 of that, which is 5.656 volts, and here it reads 5.64 volts, so in actual fact, the peak is 7.97743V - close enough for my pair of glasses!
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps
Back to Top
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Retired

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 16298
Post Options Post Options   Thanks (1) Thanks(1)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 27 Nov 2022 at 9:41am
So let's compare circuits...


Mitsubishi 1968


Me 2022

So I thank them for the practical idea, as well as all the publications during the 70's of variations on the circuit in various electronics magazines, not forgetting Tobey and Dinsdale and other versions by various manufactuers at that time.

Yes, there were faults, but SPICE modelling was still in its infancy, and most likely prohibitively expensive to run. There were no desktop computers, just those giant machines with reels of tape for memory.
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps
Back to Top
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Retired

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 16298
Post Options Post Options   Thanks (1) Thanks(1)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 27 Nov 2022 at 11:17am
Now, the way these preamps worked at "line-level" was to switch the upper and lower RIAA response out, thus leaving the 1kHz gain, and in all honestly, that's 2 or 3 dB less than 1kHz RIAA gain, so around 35dB, and that's a gain of approx. 60.

So for 400mV (revised from 500mV), the input would be 6.7mV.

Now, that requires some serious attenuation of a 2V input, but think of the noise figure input impedance range, and if we made a potential divider using a 300k resistor into a 1k resistor, the source impedance is just less that 1k, and the noise voltage is 4nV per root Hz, which is low.

So forget about the waste (lots of portable recording desks do just the same), as it has a great trade off...

It keeps adjacent switch contacts at a similar level!

So, instead of 2V on one contact, and 5mV on an adjacent contact, which would co-channel interfere like hell, it doesn't.

One thing we need to address is the 220uF capacitor in the "line-level" NFB, because as we switch between NFB networks, that capacitor will have to charge. And it's going to make a loud thump in the speakers.

Simply by wiring it out, we'll have too much DC current flowing in T1's emitter resistor, and that will change the DC output level - another loud thump!

We really need to take all the NFB networks after the output capacitor. That capacitor will be already charged to its steady state, and OK, we'll be dragging a bit more current from it, but it shouldn't make as much of a noise.

And this is where the switching needs to be make before break, to help with the sudden change.

Now, on posh preamps, a spare contact is used to drive the output relay to off for the duration of switching, but we don't have a contact spare. So those who remember "vintage" days, will remember that they turned the volume down before switching.

So, with that in mind, then it's OK.

The only problem I can see is that using the 1kHz RIAA resistor value (33k in the schematic) is that we don't have much loop-gain at 20kHz. It's probably 12dB, and the 20kHz distortion will rise, and we could do without that. So if we half the value we'll get twice the loop-gain - twice the NFB - and that should halve the 20kHz distortion.

So that would be an 18k resistor. Oh, look what the Mitsubishi uses!

But, coming after the output capacitor, the 3.3uF capacitor, the load on the output cuts the bass end, which in turn leads to bass peaking by that same amount, via NFB. In fact, it's a horrible spike, and it will clip at relatively low volumes.

All we can do is increase the value of the output capacitor to the point there is no bass peaking. Using an 18k NFB resistor, that means 100uF. And that assumes a 22k total load on the preamp output.

So now we have to think about that load.

The load on the tape out is the recorders input impedance and that was normally 50k.

The next item is the volume control, and placing it here breaks with tradition, but it might be 100k and loaded with a balance control which might be 50k, so in total we may have 20k, although having them all in parallel at max. setting would be a rather odd combination, so perhaps we go for 22k?

Thinking ahead, we could put a 50k pot in the record output, to assist with record levels if used into an ADC, but if used with a traditional cassette deck, that would be 25k full up, and the total load will be about 14k ohms.

Modelling with the 18k NFB resistor and 100uF output capacitor into a 14k load shows no peaking, but 47uF does, so it has to be 100uF.
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps
Back to Top
Sylvain View Drop Down
Senior Member
Senior Member


Joined: 18 Jan 2010
Status: Offline
Points: 481
Post Options Post Options   Thanks (0) Thanks(0)   Quote Sylvain Quote  Post ReplyReply Direct Link To This Post Posted: 27 Nov 2022 at 11:36am
The last post is very clarifying of many issues with transferring CD and Vinyl and even cassette tape to computer disc ......particularly the Load issue of Potentiometer and Attenuation of CD output...need to revise my less than good CD and Vinyl and audio Cassette transfer to computer drive. Any gratitude to You 
Back to Top
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Retired

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 16298
Post Options Post Options   Thanks (2) Thanks(2)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 27 Nov 2022 at 1:55pm
Rather than bore you with all my line-level snaps...

Alright then, I couldn't resist:


This is the LF spike I was referring to earlier. Now, at 25 milli-decibels I know it won't lead to premature clipping, and neither will you hear it. But I'm sure a certain nerds forum will sneer at it.


Signal to noise ratios better than 80dB, unweighted, and 20Hz - 20kHz are golden!


This is the THD1k including noise, remembering that the line-level input is greatly attenuated down to fit, then a lot of this is noise (but very low noise as the previous image shows).


So what we do is dig the signal out of the noise, and this required lifting the signal 7dB, and the true THD figure of better than 0.005% is revealed.


And this is the THD1k at maximum output. +23dB is excellent headroom.


This being the maximum output.


The full frequency range THD sweep shows the rising distortion I referred to earlier. 0.02% at 20kHz is pretty good!


This shows the phase flatness. It's about +/- 10 degrees 20Hz - 20kHz, and that's considered phase flat!

As a line stage, this performance is excellent.

The voltage we need to attenuate to is 12mV, and that being on an adjacent contact to the RIAA's 5mV, won't crosstalk all that much.

That none should be able to buy or sell without a smartphone and the knowledge in how to use apps
Back to Top
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Retired

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 16298
Post Options Post Options   Thanks (2) Thanks(2)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 27 Nov 2022 at 9:38pm
Now for the hard bit: the tape head!

Sub 1mV signals like MC cartridges are difficult enough, but at least they're low impedance and low inductance, and you can make the noise quite low using a transistor with a wide base.

But tape heads are high impedance, almost as high as an MM cartridge.

What's worse is their frequency response, and working with the flat output AP signal generator, what is 1mV at 1kHz is about 16mV at 50Hz, so to get it to fit without flat topping at the bass end means reducing the generator output and that puts the kibosh on noise.

Also consider the NFB EQ impedance: the bass feedback resistor is 4.7 megohms!

Measurements are difficult to untangle, and are difficult to present with any real meaning.

So, why do it? Straight away I'm going to blame Stu from HiFi Pig for planting the idea a few years ago. Whether he's still as enthusiastic about it I don't know, but at the time he was convinced that a preamp with tape head EQ was the "next big thing", and he kept talking about it for a good hour.

I've tested the 3.75 ips and 7.5 ips EQ and taken the photos of what the AP displayed, and now I've got to try and make sense of them.
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps
Back to Top
 Post Reply Post Reply Page  <1 254255256257258 345>
  Share Topic   

Forum Jump Forum Permissions View Drop Down

Forum Software by Web Wiz Forums® version 12.01
Copyright ©2001-2018 Web Wiz Ltd.

This page was generated in 0.125 seconds.